Yohannes Belayneh, Zemene Markos, Yosef Abera Email: {yohaness.belayneh, zemene.markos, yosef.abera}@ethiotelecom.et Abstract Currently the Universal Mobile Telecommunications System (UMTS) is giving services to customers which need to have an end-to-end QoS ensured. Thus there should be a design approach in the UMTS backbone network to accommodate the needs of these customers to ensure the delivery of a real time QoS across the network. This paper presents a review of how the Differentiated Service (usually referred to as DiffServ) can be integrated on the UMTS backbone network in order to ensure QoS. This integration requires mapping of different services from the UMTS network to DiffServ and addition of different hardware and software to accommodate the changes. In this paper how DiffServ can be integrated specifically for the UMTS network and how the mapping can be done will be analyzed and presented.
Key TermsQoS, UMTS, DiffServ, IP,
I.
INTRODUCTION
While considering the GSM (Global System for
Mobile Communications) in this era it is an obvious fact that the bandwidth in use is increasing drastically and there are many new applications which need high bandwidth for operation. One of the enhancements of the GSM is the General Packet Radio Service (GPRS) which accommodates these requirements. Through GPRS, GSM is evolving into the third generation (3G) cellular network called Universal Mobile Telecommunications System (UMTS). UMTS is a network- wide standard for 3G based on GSM communication standard, which provides access to a wide range of services [3]. In UMTS the structure is changed to a hierarchical structure as compared to the previous generations which gives a separate path for services which in turn ensures that UMTS is strongly service oriented. When seen from another perspective the growth of the internet is forcing networks to adapt architectures and services available in the internet network. The current trend is to have Internet Protocol (IP) as an intermediate layer for all the telecommunication services which make IP as the major transport protocol in UMTS networks with all its advantages and drawbacks. Although using IP as a transport protocol in UMTS networks has its own advantages, the lack of QoS in
networks too. Thus all the services are treated in the
same way so that there is no differentiation between services which need high priority and those that can endure greater time delays. In addition to this the integration of QoS into these IP based UMTS networks is a challenging task as the two networks are separate and are using different architectural mechanisms[1]. 1.1. QoS Quality of Service (QoS) is the mechanism insuring that a service can be delivered to the end user in an acceptable time-frame and that the service properties are stable over time within predefined boundaries [2]. Thus when seen from networking point of view, QoS can be defined as a mechanism in which some parameters of ensuring quality are defined and the network works towards keeping these parameters in range without the use of additional resources or increase in bandwidth over the network. In traditional GSM networks, as they are circuit switched networks, the QoS is highly defined as there are dedicated lines available end-to-end not only during times of transmission but also during idle times which highly decreases the efficiency of the network. In UMTS networks using packet switching technologies, QoS requirements should be handled carefully in order to address and utilize the advantages of the packet switched networks over their circuit switched counterparts. 1.2. UMTS QoS Architecture As it is discussed with QoS in different literatures, End-to-end QoS means that the evaluation of the service is done from the end-user perspective [2] [4]. The end user could be a terminal or even another 3G network. The end-to-end QoS UMTS requirement implies that QoS management is needed in all involved domains: wireless domain, IP core, external IP network. As defined by Third Generation Partnership Project (3GPP) UMTS relies on a layered Bearer Service (BS) architecture, where each layer is using the services of the layers below while providing its services to the layers above, as shown in the bearer architecture diagram of Figure. 1. According to [5], a bearer service is a service which provides the capability for data transmission between usernetwork access points. Each bearer service is defined using its QoS parameters, which define such things as traffic type, bit rates, error ratio.
2.
Figure 1: UMTS QoS Architecture
The entities involved the UMTS packet data architecture which provides QoS services are the following [2]: The User Equipment (UE) is the UMTS terminal. Mobile Terminal (Mobile Terminal) and Terminal Equipment (TE) are the functional blocks of the UE. MT is the part which interfaces with the wireless domain, being responsible for handling of radio signals. TE provides the user interface and the application. The Radio Access Network (RAN) provides the interface to the core network and can be composed of Base Stations (the UMTS evolution of the GSM Base Transceiver Station) and the Radio Network Controllers (the UMTS evolution of the GSM Base Station Controller). The BSs task is to transmit data over the radio interface to and from the MT and to interface with the RNCs. The Core Network Edge Node (CN Edge Node) provides the interface between the RAN and the packet switched core assuring session and mobility management.
The CN Gateway connects the packet
switched core network to external networks. It maintains the routing information necessary to tunnel the packet data between the networks, has mostly functions typical for gateways in IP networks. 1.3. DiffServ Differentiated Services (DiffServ) is an IP-based QoS technology which attempts to solve the QoS problems without making any resource reservation for flow establishment and
maintenance [4]. The control is based on
DiffServ Code Point (DSCP) which is contained in the IP header which dictates the routers on how to handle traffic differentiation. The main advantage of the DiffServ architecture is its simplicity. On the other hand, as all the flows with the same DSCP value are seen as a single flow with an unique per-hop behavior (PHB) DiffServ provides just relative QoS, in the sense that that the QoS for one flow depends on the behavior of other flows. UMTS QoS Requirements The UMTS QoS general requirements define the constraints the set of attributes characterizing the QoS should meet. The QoS mechanisms have to allow efficient use of radio capacity and efficient resource utilization It should allow independent evolution of core and access networks The UMTS network should be evolved with minimized impact on the evolution of transport technologies in the wire line networks The overhead and additional complexity caused by the QoS scheme should be kept reasonably low, so as the amount of state information transmitted and stored in the network The QoS behavior should be dynamic, i.e., it should be possible to modify QoS attributes during an active session
When defining the UMTS QoS classes the
restrictions and limitations of the air interface have to be taken into account. The QoS mechanisms provided in the cellular network have to be robust and capable of providing reasonable QoS resolution. There are four different QoS classes: - Conversational class - Streaming class - Interactive class and - Background class. The main distinguishing factor between these QoS classes is how delay sensitive the traffic is: Conversational class is meant for traffic which is very delay sensitive while Background class is the most delay insensitive traffic class. Conversational and Streaming classes are mainly intended to be used to carry real-time traffic flows. A table showing the QoS classes for UMTS is shown below: Table 1: UMTS QoS Classes
In accordance with this, the main challenges that the
UMTS QoS architecture has to overcome are: Translation parameters and mechanisms -service differentiation based on a set of traffic classes needs a simple and reliable translation mechanism between the different domains involved. UMTS QoS Management - The network should be monitored and managed to assure the implementation of the user agreements. Negotiation and modification of the QoS available from the network should be possible. End-to-end QoS has two dimensions: a vertical one which refers to the mapping of high level bearer service attributes into lower level bearer service parameters and a horizontal one which implies translation of the QoS attributes and QoS management mechanism between the domains. Table 2 presents the UMTS Bearer Service parameters for each class. Table 2: QoS attributes in UMTS Bearer Service
2.1. Network level QoS Translation management on
the wireless interface The control plane structure for and end-to-end IP QoS management is shown in figure 2. As can be seen, to manage the end-to-end QoS there are two extra components in the control plane: IP BS (Bearer Service) Manager and P-CSCF (Proxy Call State Control Function). The IP BS manager is used to control the external IP bearer service. To interact with UMTS Bearer
Service, the IP BS Manager leverages the Translation
Function to map the mechanisms and parameters used within the IP bearer service to those used within the UMTS bearer service. As can be seen there are two IP BS Managers; one in the UE (User Equipment) one in the Gateway (the Gateway might be a GGSN). The IP BS Managers in the UE and the GGSN could communicate with each other using relevant signaling protocols. The P-CSCF is a mobiles first contact point for IP multimedia sessions. It is essentially a local SIP server which includes Policy Control Function (PCF). The PCF is a logical entity for policy decision, which conforms to the policy framework defined by IETF and coordinates the applications with the resource management in IP layer. In the UE the application layer requirements are translated into the IP Bearer Service requirements by using application specific mechanisms. By using the Translation function the IP BS Manager translates the IP specific QoS mechanisms into the PDP Context properties. By using the service of the underlying bearer service managers the UMTS BS Manager sends the PDP Context request to the UMTS BS Manager serving the Gateway. As the external network is an IP based network the PDP Context information is translated back into the IP Bearer Service properties using the Translator functionality residing in the Gateway. The Translator functions need information about mapping between the UMTS and the external IP QoS.
Figure 2: Control plane for end-to-end IP QoS
management 2.2. Network level QoS Translation management on the external network interface If the external IP network is DiffServ enabled then the Translator should be connected to a Service Level Agreement (SLA) repository. The SLA is a contract between the provider of a service (in this case the external IP network) and the customer (the UMTS
network) which specifies the services offered to the
customer. Both edge and core DiffServ routers use to be supported in order to implement DiffServ. The DiffServ mechanism is scalable and it can be implemented with long term setup. With a special marking in the packets IP header based on DiffServ code points, it is possible for routers inside the network to process the packet flow. The packets with the same code points operate with forwarding behavior PHB (Per-Hop Behavior) inside the domain. DiffServ defines three classes: expedited forwarding (EF), assured forwarding (AF) and best effort. Figure 3 shows the mapping of the DiffServ classes to the UMTS QoS classes depending on the type of service class. The advantage of this mapping is the absence of extra signaling along the path.
groups according to predefined rules. The Traffic
Marker sets the value of the DSCP (Differentiated Services Code Point) field so that the packet receives appropriate PHB. The Traffic Meter measures the rate of traffic streams selected by the classifier and Traffic Sharper may drop packets if there is not sufficient buffer space. The traffic classes, traffic policies and associated Access Control List (ACL) are configured by the edge routers. 3. Conclusion This paper summarizes the concept behind using DiffServ with UMTS networks in order to ensure end-to-end QoS requirements which is the requirement encompassing the path from the sending UE to receiving UE. This can be achieved by the integration of the IP network with DiffServ routers enabled and mapping the traffic classes in the UMTS QoS to the DiffServ classes. The points summarized in this paper can be verified using simulations and case studies which are beyond the scope of this paper. REFERENCES
Figure 3: Mapping of DiffServ classes to UMTS QoS
classes The DiffServ routers do not have to store any state information of the flows. These routers consist of following components: Packet Classifier, Traffic Marker, Traffic Meter and Traffic Sharper. The Packet Classifier performs the identification of incoming packets and their separation into several
[1] Robert Lloyd-Evans. QoS in integrated 3G Networks.
Artech House, 2002 [2] Claudiu Chioariu. QoS in UMTS. Helsinki University of Technology Seminar on Internetworking 2004, T -110.551 [3] 3GPP TS 22.060 v6.0.0. Technical Specification Group Services and System Aspects; General Packet Radio Service (GPRS); Service Description, Stage 1 [4] 3GPP TS 23.107 v6.0.0. Technical Specification Group Services and System Aspects; Quality of Service (QoS) concept and architecture [5] Timo Halonen, Javier Romero and Juan Melero. GSM, GPRS and EDGE Performance. Wiley, 2002